GH Issues Pilot open

[R.A. Porter]

Good news, everyone!

The GH Issues Pilot has started. For the next few weeks, when creating new issues please endeavor to create them as GH Issues (https://github.com/codice/ddf/issues) instead of in Jira. Create them with the same attention to quality and detail that you have always applied. For the small differences, please check here (https://github.com/codice/ddf/wiki/%5BGH-ISSUES-PILOT%5D-Creating-Issues).

PRs created against GH Issues will use the "Fixes: #____" functionality to link PR to issue; this has been added to the PR Template for you.

Also note that there is a sample kanban board that is also being tested during the pilot program (https://github.com/codice/ddf/projects/3) but that will be monitored and managed by a handful of committers as an experiment. Feel free to view it but don't worry about doing anything with it.

One last note...because of the risk of issue number overlap, we're going to request a small change to your commit message prefix.

For commits created to resolve GH Issues during the pilot and beyond, please prefix the issue number with a 0 until we roll over at issue #10000, ie: DDF-0####. This kludgy solution will avoid overlap with issues created in Jira.

Thanks!

-Richard

[Shaun Morris]

Does anybody have experience with using any boards with GH Issues? I've seen Waffle and Glo Boards. We could also experiment with some board during this pilot period.

[R.A. Porter]

We've been experimenting with GH projects - all kanbans so far - on the pilot but it's still very much a WIP. Right now, there are three of them here (https://github.com/codice/ddf/projects).

Two - the documentation and "GH Issues Pilot Test" - are kanbans with basic automated workflows; the third is a kanban with no triggers. The pilot test board is pretty useless, as it's just a catchall where I've been adding any new issues/PRs that do not specifically fall into the other two boards, but I wanted to see if it brought any value even when that ill-defined. The other two boards seem like they might be more useful, narrowing their focus to specific topic areas. Ideally, we'd probably want to move toward more epic-based boards, with a handful of catchalls, and let the engineers working those epics decide how they want to manage them and structure them.

Conversely, GH Projects are still too limiting and we'll find them not that useful.

-Richard